Table 5 Experimental classification results
From: Power equipment fault information acquisition system based on Internet of things
Algorithm | Classification | Accuracy (%) | Recall (%) | F1 value (%) |
|---|---|---|---|---|
KNN | High-risk | 78.2 | 81.6 | 75.7 |
Low risk | 64.8 | 63.6 | ||
Invalid | 79.0 | 82.7 | ||
Crack | 78.6 | 82.9 | ||
SVM | High risk | 86.6 | 87.8 | 86.3 |
Low risk | 89.2 | 88.6 | ||
Invalid | 86.8 | 90.4 | ||
Crack | 79.8 | 80.5 | ||
IOT approach | High-risk | 98.0 | 97.3 | 96.4 |
Low risk | 94.9 | 96.6 | ||
Invalid | 97.9 | 95.2 | ||
Crack | 95.4 | 92.8 |